© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 31101
    Prev
    Next
    INDEX
    Explanations

    high-frequency numerical values or counts in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    18
    +0.12
    0.7%
    499
    +0.12
    0.7%
    165
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    165
    +0.12
    0.03
    18
    +0.12
    0.03
    273
    +0.12
    0.02
    Negative Logits
    ĻĤ
    -2.26
    ľĵ
    -1.89
    ı
    -1.84
    »¿
    -1.81
     <-
    -1.78
    Ĺ
    -1.74
    ħ
    -1.74
    ypse
    -1.69
    Ĭ
    -1.67
    ¦
    -1.64
    POSITIVE LOGITS
    qualified
    1.70
    \[[@
    1.53
    Univers
    1.53
    [*
    1.51
    -*
    1.50
    funded
    1.50
    Foundation
    1.50
    developed
    1.45
     qualified
    1.45
    represented
    1.42
    Activations Density 0.111%

    No Known Activations